Reading :: The promise and peril of entrepreneurship

 The Promise and Peril of Entrepreneurship: Job Creation and Survival among US Startups 

By Robert W. Fairlie, Zachary Kroff, Javier Miranda, and Nikolas Zolas

In this book, written by economists who study startup job creation, the authors report results based on their dataset, the Comprehensive Startup Panel: A dataset that tracks job creation and startup survival. The authors created this new dataset because current Census Bureau data don’t do a good job of tracking startups. Although startups may eventually become employers, most startups begin life as nonemployer firms, so they had to reach across data for both nonemployer firms (NEFs) and employer firms to create a more comprehensive picture of startups (p.2). They investigated three questions:

• “How many jobs are created by the average entrepreneur?” (p.2) — and, related, “is entrepreneurship about creating jobs or about creating a job?” (p.3). 
• “Do these created jobs last or do they disappear quickly?”
• “How many entrepreneurial firms survive each year after startup?” (p.2). 

“The US Bureau of Labor Statistics (BLS) defines entrepreneurs as new employer establishments” (p.3), but NEFs represent both the majority of startups and the majority of total firms in the US (pp.3-4). Once the authors account for this by creating a new dataset, one that unites NEFs and EFs and defines a business as an entity that has revenue tax liability (p.16), they find that 

• After one year, 59% of startups survive
• After two years, 47%
• After 5 years, 33% (p.11; p.65)
• After 7 years, only 29% (p.65)

compared to the official statistics of EFs, which claim a 50% survival rate after five years — a figure that ignores startups’ previous existence as NEFs (p.27).

Many startups also have quick exits:

• Year 0, any given year, starts with about 4.1m startups 
• Year 1, only 2.4m of these survive
• Year 2, only 1.9m survive
• Year 5, 1.2m survive (p.63)

Including NEFs is important because nonemployer businesses have a startup rate of 35%, which is three times that of employer businesses (p.97). 

They use a broad definition of entrepreneur: Did someone create a job for themselves (p.100)? 

Finally, they examine startup owners: Who owns startups? Their statistics here reflect what we already know about startup owners: They are disproportionately white and male, with a college degree (p.143). 

Overall, I really appreciated this book. I am no economist, but I appreciated the authors’ clear methodological assumptions and their accounting for how they got the numbers they did. The results helped me to better understand the startup landscape. If you’re similarly interested in startups, definitely check it out.